Predicting The Future Through The Zeitgeist

Read a fascinating piece by my friend Ulrike Reinhard, an interview with Peter Kruse, the founder of nextpractice. He has worked to create a method to predict societal change by tapping into the collective insight embedded in the minds of groups of experts in various fields.

For example, if were interested in trends in food fashion — like Chipotle might have been before concocting their newest chain of Asian restaurants — Kruse would subject a small group of foodistas to his techniques, and, voila!, confirm that Americans are ready to eat Vietnamese noodles.

Ulrike Reinhard, Soft Values, Hard Facts

Reinhard: The mantra of nextpractice is ‘A Matter of Fact in a World of Values’. What does this mean?

Kruse: In established methods of collecting data, like standardised questionnaires and predefined scales, people give their judgments on the basis of hopefully intelligent questions and simple categories like ‘yes’ and ‘no’, multiple choice, ranking, etc. The respondent can only add value when the intentions of the interviewer are decoded correctly.

But language is a very tricky phenomenon, so the first difficulty to be tackled is the problem of semantics, which adds a lot of noise to every measurement. The second problem is a direct consequence of the first. Interpretation of language is a mainly conscious process that isn’t well connected with a person’s intuitive knowledge and unconscious valuations, which are crucial for complexity reduction.

Only when people are given total freedom to explain something in their own words – as in a qualitative interview – can their full potential to add reasonable information be enabled. But a qualitative interview only shifts the problem of semantics over to the person collecting the data. Now the one listening to the answers is in charge of interpretation.

To solve this dilemma of quantitative versus qualitative measurement, about 20 years ago we started a project to develop an interview technique that combines the strengths of both forms of measurement. We aimed for a method that was able to get full access to a person’s unconscious valuations and was then capable of mathematically combining this individual data into a common picture that merited the name of ‘collective intuition’. Our nextexpertizer computer-based interview tool is the result of all these endeavours.

What it basically does is, first, creates a list of elements for comparison. This list can contain up to about 60 short word descriptions, pictures or even video clips to direct the attention of a person to a chosen topic of interest.

Then a relatively small sample of 100-200 people with in-depth practical experience of the topic of interest is interviewed. The interview is strictly ritualised. Interviewees are confronted with two elements from the predefined list, asked to decide whether these two elements are more similar or more different, and to describe in their own words why they are similar or different. The ‘experts of experience’ are then told to rate all the other elements very quickly on the basis of the dimension they have created for differentiation.

With every decision, a slight indication of their unconscious value system passes the threshold.

At the end of the procedure, which usually takes up to two hours, every interviewee has described his or her picture of the topic of interest in a matrix associating all the elements of comparison by 10-20 freely formulated dimensions. Due to the enormous amount of decisions taken to create the matrix, there is no chance of intellectual control. The meaning of comments can be assigned by the way words are used to define the relationship between the compared elements. By calculating three-dimensional representations of the vector spaces created by the expert interviewees, it is easy to perceive directly the implicit value system they share. The right data is presented in the right way for the upgrading of decision-making processes. Unconscious soft data is transformed into facts and mathematically defined key performance indicators.

The measurement of collective intuition is a very promising alternative for understanding the actual behaviour and predicting the future behaviour of customers, citizens and other persons involved in complex cultural order formation processes. As our studies show: you can be years ahead.

I can see this is social network terms: finding a small subset of very well-connected experts in some domain (and by well-connected, I mean in terms of betweenness, not popularity), and expose them to a collection of concepts that have been whittled down to their most atomic level. The group can act like the sensor in a camera, each receiving and passing along one pixle of the resulting image, and the collective intuition of the group yields a picture of where the future is moving.

Apparently, Kruse has worked with companies like Volkswagon:

Kruse: For Volkswagen, for instance, we analysed the cultural value system for cars in Europe and Asia since 2006. Years before the financial crisis – and clearly ahead of the decline in the premium segment in Germany and other mature markets – changes in cultural value systems indicated a significant breakdown in the status function of cars. People’s preferences turned to the functional aspects of mobility – a big chance for public transportation. The collective intuition of a few hundred people was able to anticipate an upcoming shift in consumer behaviour that went against the grain of what public opinion and the mass media were saying. Even the growing importance of the new segment of small premium cars was indicated long before any real increase in sales volume.

So, can prediction become commonplace? If futurism just a social process? Can we simply crank the handle and have better insight into what’s about to transpire, like a macroeconomic version of Minority Report? Can Peter Kruse really see the future though our eyes?

I am going to have to look more deeply into Kruse’s work. More to follow.